Run a FLORA technique
Hosted MCP generation inputs are plural: call flora_create_generations with { "generations": [{ "workspace_id": "ws_…", "project_id": "prj_…", "type": "image", "prompt": "…" }] } (1–20 items). Put per-generation fields, including optional model, params, and reference_node_ids, inside each item. Read generations[] in the response; retain successful entries' run_id and handle failures individually. Poll flora_list_generations with { "run_ids": ["run_…"] }, even for one run; add technique_id for technique runs. Never retry successful items because another item failed.
Use dedicated tools for this workflow, including batches. execute is deprecated; use it only for an SDK operation without a dedicated tool. SDK examples below describe orchestration: use the corresponding dedicated tools, issue independent calls concurrently, retain every run id, and poll in later calls. Do not choose execute just to combine calls.
A technique is a saved workflow that chains several models. It is the thing FLORA
does that a single image model cannot: the steps, prompts, and model choices are
fixed by whoever authored it, so the same technique gives the same treatment every
time.
Input: the user's intent, plus any source image.
Output: the technique's output URLs, the charged cost, and the project the run
landed in.
Steps
Resolve the workspace. Call flora_list_workspaces. If there is more than
one, ask which to bill before spending anything.
Find the technique. Call flora_list_techniques and match on the user's
intent. It returns run_cost and a summary of inputs for each. If nothing fits,
say so and offer flora_create_generations instead — do not force an unrelated technique.
Get the exact input ids. Call flora_get_technique. Its declared input ids
are the keys flora_run_technique expects. Never guess them from the name.
Get each image input to an HTTPS URL (see Image inputs below).
Run it. Call flora_run_technique with workspace_id, technique_id, and
an inputs object keyed by those ids. Omit optional text inputs entirely when
the user gave no direction — an empty string is rejected, not treated as absent.
Poll and report. Poll flora_list_generations until status is completed or
failed. Report the output URL, the charged cost, and a link to the project.
Image inputs
FLORA fetches images server-side from an HTTPS URL. Files the user attaches in
ChatGPT already have one — they are hosted on files.openai.com or
cdn.openai.com, both allowlisted. Pass that URL straight to
flora_create_asset as source, or into the technique input directly.
Never base64-encode a file or try to upload bytes. If the image genuinely has no
URL, say that this environment cannot upload it and point the user at their FLORA
project to add it there.
Rules
- State the cost before running.
run_cost comes back from
flora_list_techniques. Techniques range from free to several dollars per run.
- You cannot see the output. Runs return URLs, not pixels. Describe what was
produced from the technique and the inputs; never claim to have looked at it.
- On
input_validation_error, re-read flora_get_technique, rebuild inputs
from what it returns, and retry once. Do not retry other failures — retries bill.
- On insufficient credits, surface the message and stop.
1---2name: flora-run-technique3description: Run a saved FLORA technique — a multi-step visual workflow such as background swap, relighting, upscale, model swap, or sketch-to-render. Use when the user names a technique, asks to transform an image they already have ("put this on a clean background", "match this lighting", "upscale this", "make this look photographed"), or asks what FLORA can do. Do not use for a plain text-to-image request with no source image, and do not use to inspect existing project work.4---56# Run a FLORA technique78Hosted MCP generation inputs are plural: call `flora_create_generations` with `{ "generations": [{ "workspace_id": "ws_…", "project_id": "prj_…", "type": "image", "prompt": "…" }] }` (1–20 items). Put per-generation fields, including optional `model`, `params`, and `reference_node_ids`, inside each item. Read `generations[]` in the response; retain successful entries' `run_id` and handle failures individually. Poll `flora_list_generations` with `{ "run_ids": ["run_…"] }`, even for one run; add `technique_id` for technique runs. Never retry successful items because another item failed.910Use dedicated tools for this workflow, including batches. `execute` is deprecated; use it only for an SDK operation without a dedicated tool. SDK examples below describe orchestration: use the corresponding dedicated tools, issue independent calls concurrently, retain every run id, and poll in later calls. Do not choose `execute` just to combine calls.1112A technique is a saved workflow that chains several models. It is the thing FLORA13does that a single image model cannot: the steps, prompts, and model choices are14fixed by whoever authored it, so the same technique gives the same treatment every15time.1617**Input:** the user's intent, plus any source image.18**Output:** the technique's output URLs, the charged cost, and the project the run19landed in.2021## Steps22231. **Resolve the workspace.** Call `flora_list_workspaces`. If there is more than24 one, ask which to bill before spending anything.25262. **Find the technique.** Call `flora_list_techniques` and match on the user's27 intent. It returns `run_cost` and a summary of inputs for each. If nothing fits,28 say so and offer `flora_create_generations` instead — do not force an unrelated technique.29303. **Get the exact input ids.** Call `flora_get_technique`. Its declared input ids31 are the keys `flora_run_technique` expects. Never guess them from the name.32334. **Get each image input to an HTTPS URL** (see Image inputs below).34355. **Run it.** Call `flora_run_technique` with `workspace_id`, `technique_id`, and36 an `inputs` object keyed by those ids. Omit optional text inputs entirely when37 the user gave no direction — an empty string is rejected, not treated as absent.38396. **Poll and report.** Poll `flora_list_generations` until status is `completed` or40 `failed`. Report the output URL, the charged cost, and a link to the project.4142## Image inputs4344FLORA fetches images server-side from an HTTPS URL. **Files the user attaches in45ChatGPT already have one** — they are hosted on `files.openai.com` or46`cdn.openai.com`, both allowlisted. Pass that URL straight to47`flora_create_asset` as `source`, or into the technique input directly.4849Never base64-encode a file or try to upload bytes. If the image genuinely has no50URL, say that this environment cannot upload it and point the user at their FLORA51project to add it there.5253## Rules5455- **State the cost before running.** `run_cost` comes back from56 `flora_list_techniques`. Techniques range from free to several dollars per run.57- **You cannot see the output.** Runs return URLs, not pixels. Describe what was58 produced from the technique and the inputs; never claim to have looked at it.59- **On `input_validation_error`,** re-read `flora_get_technique`, rebuild `inputs`60 from what it returns, and retry once. Do not retry other failures — retries bill.61- **On insufficient credits,** surface the message and stop.
florafauna-ai